OmicsBox Engine Pipelines
Introduction
The OmicsBox Engine pipelines are ready-made Nextflow and Snakemake workflows for complete analyses, such as assembling and annotating a genome or testing for differential expression. Each step of a pipeline is one OmicsBox Engine CLI command.
The work is split between the 2:
- OmicsBox runs the analyses in the cloud and handles parallelization, resources and input validation.
- The workflow manager runs the steps in order, passes each step's outputs to the next one and organizes the results.
The pipelines are open source, under the Apache License 2.0, in the omicsbox-engine-workflows repository. You can use and modify them freely. Running them requires the OmicsBox Engine CLI and an OmicsBox subscription with the Engine add-on.
Requirements
- The OmicsBox Engine CLI, installed and on the
PATHof the machine that runs the pipeline steps. - A signed-in CLI on that machine, with
omicsbox author theOMICSBOX_API_KEYenvironment variable. See Installation and Sign-in. The pipelines take no credentials of their own. - Nextflow 25.04.0 or later, or Snakemake 9.23.0 or later.
If the CLI is not signed in, the first step of the pipeline fails with a clear message, and no cloud analysis is submitted.
Available pipelines
| Pipeline | What it does | Nextflow | Snakemake |
|---|---|---|---|
| De Novo Transcriptome Characterization | Assembles a transcriptome from RNA-Seq reads without a reference genome, assesses it, predicts coding regions and annotates the proteins. | ✓ | |
| Eukaryotic Genome Analysis | Assembles a eukaryotic genome from short reads, evaluates and masks it, predicts genes and annotates them. | ✓ | |
| Functional Annotation | Assigns Gene Ontology terms, protein domains, orthology groups and enzyme codes to a set of transcripts, coding sequences or proteins. | ✓ | ✓ |
| Gene Level Analysis | Quantifies gene expression from short-read RNA-Seq against a reference genome and annotation, then tests for differential expression. | ✓ | |
| Genetic Variation Analysis | Calls, filters, phases and annotates variants from short-read DNA sequencing, and tests them for association with phenotypic traits. | ✓ | |
| Long Reads Eukaryotic Genome Analysis | Assembles a eukaryotic genome from PacBio or Oxford Nanopore reads, polishes it with short reads, then predicts and annotates genes. | ✓ | |
| Long Reads Prokaryotic Genome Analysis | Assembles a bacterial or archaeal genome from PacBio or Oxford Nanopore reads, polishes it with short reads, then predicts and annotates genes. | ✓ | |
| Metagenomics Analysis | Characterizes the genes and functions of a microbial community from shotgun reads, removing host and contaminant reads before assembly. | ✓ | |
| Prokaryotic Genome Analysis | Assembles a prokaryotic genome from short reads, optionally with mate-pair or hybrid data, and annotates the predicted genes. | ✓ | |
| Transcript Level Analysis | Quantifies isoform expression from short-read RNA-Seq against a reference transcriptome, then tests for differential expression. | ✓ |
Snakemake versions of more pipelines are in preparation. Each pipeline's step-by-step description is in the repository wiki.
How a pipeline is configured
Every pipeline comes with a configuration template: a single file that lists all its settings, with the accepted values and the default of each one written next to it. You do not need to learn the CLI syntax to configure a pipeline.
Running a pipeline always takes 3 steps:
- Export the template to your working folder. This only copies the file, and nothing is sent to the cloud.
- Edit your copy: set the input files and the output folder, and enable or change the options you need.
- Run the pipeline with your copy of the template.
Keep your edited copy: it records exactly how the run was configured.
The exact commands for each workflow manager are on the Nextflow and Snakemake pages.